Agri-environmental schemes (AES) are a central component of the environmental policy of the European Union. Despite widespread interest and investment in AESs, few of these pro- grams have been carefully evaluated and doubts are often expressed about the effectiveness of voluntary programs. The purpose of this article is to estimate the additional effects of AESs targeting nonpoint source pollution from pesticides, focusing on one emblematic case study: herbicide use in vineyards. We use original data collected from winegrowers participating in AESs in the south of France, and we use exogenous variation in the timing of the implemen- tation of the AESs as a natural experiment. We show that the quantity of herbicides used by participants in the programin 2012 was around 30%belowwhat theywould have usedwithout the program, while the impact was significantly higher in 2011 - around 50% - presumably be- cause of higher weed pressure. Although significant, these impacts remain smaller than what had been expected by policymakers. Focusing on the “zero herbicide between the vine rows” option, which is both the most often chosen as well as the least stringent among the mea- sures,we moreover showthe presence of windfall effects. Simple extrapolation of these results suggests that this level of effectiveness may not be sufficient to ensurewater quality in thewa- tersheds targeted by the AES.